Ecological zone construction and multi-scenario simulation in Western China combining landscape ecological risk and ecosystem service value

Abstract

The disparity between eastern and western China is significant, with considerable imbalances in city development. Ecological zoning can promote the comprehensive and coordinated urbanization of China and provide valuable planning references for the future development of western cities. This study focused on Hohhot, a typical western city, to analyze the spatiotemporal evolution of its ecosystem service value (ESV) and landscape ecological risk (LER) from 2000 to 2020 using value-equivalence and landscape ecology methods. The Z-score method was employed to delineate ecological zones, and the PLUS model was used to predict the ecological zoning patterns under four scenarios for 2040. The study’s findings revealed the following: (1) During the three periods from 2000 to 2020, Hohhot City was mainly characterized by the dominance of very low risk, low risk, and medium risk ecological levels, The areas of these levels accounted for 94.11%, 91.11%, and 90.95% of the city’s total area respectively. The overall variation in ESV was minimal, with grassland, water area, forests, and arable lands constituting the primary contributors to ESV. (2) Across the four future scenarios, LER and ESV exhibited divergent trends. Under the urban development scenario, high risk areas increased the most (4.14%), while in the arable land protection scenario, very low risk areas were smallest, and low risk areas were largest. Implementing reasonable urban planning policies can enhance ESV, with notable increases in low value, medium value, and higher value areas under ecological protection, natural development, and arable land protection scenarios. (3) Combining ESV and LER dimensions, Hohhot was categorized into four ecological zones: ecological restoration reserve (Zone I), ecological rich reserve (zone II), ecological balanced protected areas (Zone III), and ecological challenge reserve (Zone IV). Differentiated management strategies were proposed for each zone. This study integrated LER and ESV for ecological zoning, offering a novel perspective for understanding ecological security. By extending temporal scales using the PLUS model, it predicted the spatial patterns of ecological zones under four scenarios, achieving dynamic ecological zoning. The findings enriched ecological zoning methodologies and provided a scientific basis for dynamic monitoring and management of ecological security.

Introduction

With the rapid global population growth, accelerated economic development, and urbanization, ecosystems and landscape patterns are facing unprecedented pressures. Human activities have increasingly disrupted the natural environment, leading to the continuous degradation of ecosystem services and heightened landscape ecological risks¹. Addressing how to implement effective ecological management measures to enhance ecosystem services, mitigate ecological risks, and achieve sustainable development has become a pressing scientific issue². Ecological zoning is regarded as a vital tool for resolving this challenge. By rationally partitioning ecosystem services and protecting ecologically fragile areas, ecological zoning groups regions with similar characteristics based on the fundamental attributes and features of ecosystems at various spatial scales, facilitating better management and protection of ecosystem functions³.

The concept of ecological zoning originated from the zoning theories of geography. Early ecological zoning primarily focused on geographical scales, emphasizing the rational development and utilization of natural environments for human purposes. It concentrated on agricultural ecological zoning by considering factors such as climate, soil, geology, and hydrology, yet lacked an understanding of the inherent patterns and dynamic changes in natural environments and ecosystems⁴. As ecological research deepened, the concept of ecological zoning evolved. In 1976, Bailey proposed the first true ecological zoning concept, defining it as a process of spatially integrating natural units from the perspective of ecosystems⁵. This theory sparked discussions among ecologists worldwide about the principles, criteria, indicators, levels, and methods of ecological zoning. However, due to differences in research objects and perspectives, the academic community has yet to establish a unified method for ecological zoning⁶. Existing ecological zoning methods can be broadly categorized into two types: single-indicator evaluations and multi-dimensional indicator evaluations. Compared to single-indicator evaluations, multi-dimensional methods integrate various factors into a unified assessment framework, reflecting changes in natural ecological environments more comprehensively⁷. This comprehensive approach ensures that ecological zoning results are more scientific and holistic. For instance, Han⁸conducted sensitivity analysis and ecosystem service importance evaluations for Qingdao, identifying spatial distribution patterns of ecological sensitivity and service importance. Based on these findings, ecological functional zoning was performed, and key priorities, goals, and measures for ecological protection and construction in each functional zone were proposed. Similarly, Chen et al. adopted a two-tier zoning system, dividing China into 12 ecosystem productivity regions based on terrain and climatic characteristics and further into 64 ecosystem productivity zones by considering arable land, forest, and grassland ecosystem types, their combinations, and productivity levels⁷. Zheng et al. used a three-tier system based on thermal zones, aridity zones, and natural regions, dividing China into 11 thermal zones, 21 aridity zones, and 49 natural regions. In summary, researchers at home and abroad have employed diverse indicators to conduct ecological zoning studies, yielding substantial results⁷. These studies have laid a solid foundation for improving ecological zoning theories and methods and provided scientific guidance for ecological protection and sustainable development strategies.

In recent years, population growth, economic development, and urban expansion have significantly impacted ecosystems and landscape distribution patterns, leading to the continuous degradation of ecosystem services and increasing landscape ecological risks^9,10,11. In response to the call for ecological civilization, various regions have actively undertaken ecosystem service evaluations, ecological risk assessments, and ecological management zoning to promote green and sustainable development. ESV and LER are key indicators reflecting the positive and negative aspects of ecological environmental quality. These indicators exhibit spatiotemporal interactions and represent two critical dimensions of ecological security assessment. Current research has primarily focused on the correlation between ESV and LER. For instance, Yan et al.¹² analyzed the relationship between ESV and LER in Wuhu City and found significant spatial autocorrelation, suggesting that future development planning in Wuhu should prioritize the conservation of wetlands and forest landscapes to enhance ESV and reduce LER. Similarly, Bian et al.¹³ conducted a nationwide analysis of China and revealed a significant negative correlation between ESV and LER. Zhang et al.¹⁴ explored the spatiotemporal evolution of ESV and LER in Xi’an, concluding that ESV is negatively correlated with the landscape ecological risk index. While these studies have thoroughly examined the relationship between ESV and LER, only a few have integrated these indicators to delineate ecological zones.

Moreover, existing research on ecological zoning has primarily focused on static delineation¹⁵. Given that ecological construction is crucial to the future well-being of humanity, it has become necessary to extend the temporal dimension of current studies, simulate future ecological conditions, and uncover spatial patterns of ecological security under different scenarios. In recent years, advancements in modeling have enabled accurate predictions and simulations of future ecological environments. Models such as SD¹⁶, FLUS¹⁷, the logistic-CA¹⁸model, and the PLUS¹⁹have rapidly evolved, allowing researchers to simulate land-use spatial distribution under multiple scenarios and further conduct dynamic ecological zoning^20,21,22.

The western region of China is considered an underdeveloped area in the total socio-economic context. This disparity between the western and eastern regions is largely attributed to unique geographic and historical factors²³. To advance China’s urbanization in a comprehensive and coordinated manner, it is imperative to establish a rational urban development plan for the Western region. Hohhot City, situated as the political, economic, and cultural epicenter of the Inner Mongolia Autonomous Region, serves as a prime example of a rapidly developing city in western China. Moreover, it plays a pivotal role as the central city within the Hubao-Elm group and stands at the crossroads of three major strategic initiatives: Northwest Development, Bohai Economic Circle, and the Revitalization of the Old Northeast Industrial Base²⁴. Currently, Hohhot is experiencing swift urban development. However, this rapid urbanization has led to a host of ecological and environmental challenges. Despite significant efforts by various institutions and researchers to address these environmental issues, there is room for improvement. Thus, this study centers on Hohhot as its research area to redefine ecological zones through the use of ESV and the LER. Therefore, this study selected Hohhot City as the research area to investigate. By integrating ESV and LER, ecological zones were constructed. Using the PLUS model with an extended temporal scale, the study predicted the spatial patterns of ecological zones under four scenarios, achieving dynamic delineation of the zones.

The primary objectives of this study are as follows: (1) to reveal the spatial and temporal patterns of ESV and LER in Hohhot from 2000 to 2020; (2) to delineate ecological zones by integrating ESV and LER and (3) to simulate the spatial patterns of ecological zones under multiple future scenarios using the PLUS model. The findings provide effective recommendations for land use planning and sustainable development in Hohhot.

Study area and data

Study area

Hohhot is situated in the central region of the Inner Mongolia Autonomous Region, bordered by the Yin Shan Mountain Range to the north and the Yellow River to the south (Fig. 1). The topography is generally lower in the southwest and higher in the northeast, with altitudes ranging from 900 m to 1200 m. Hohhot experiences a temperate continental monsoon climate with distinct seasonal changes. The city has frigid winters with a minimum temperature of −32.0℃ and hot summers with a maximum temperature of 37.3 ℃. It also exhibits significant annual temperature variations and uneven rainfall distribution²⁵.

Fig. 1

Location map of Hohhot. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA). The original boundary and data were obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Data acquisition and treatment

This study utilized land use, natural, and socio-economic factors data. The land use dataset was sourced from the Resource and Environment Science and Data Center (https://www.resdc.cn/). It primarily utilized Landsat-MSS, Landsat-TM/ETM, and Landsat 8 remote sensing image data. The data accuracy was 30 m. In this paper, the land use data were classified into six primary land categories: arable land, grassland, forest land, water areas, construction land, and unused land, following the classification method established by the Resource and Environment Science and Data Center. The natural factor data include DEM, slope, aspect, and soil type. DEM and soil type data were sourced from the Resource and Environment Science and Data Center. Slope and aspect were derived from DEM data after being processed with ArcGIS 10.8. Socio-economic factors comprised population density, gross domestic product (GDP), and distances from water areas, citys, primary roads, railroads, bus stations, secondary roads, rivers, and more. The specific data information is shown in Table 1.

Table 1 Data information.

Methods

Figure 2 illustrates the workflow of this study. We evaluated the accuracy of the GeoSOS-FLUS and PLUS models. Subsequently, we identified the most suitable land use simulation model for Hohhot. This selected model was then integrated with ordinary kriging interpolation and Fragstats 4.2 (http://www.umass.edu/landeco/research/fragstats/fragstats.html) within ArcGIS to investigate changes in land use, the spatial distribution of LER, and ESV patterns in Hohhot. Additionally, we forecasted the future trends in land use, LER, and ESV for Hohhot, shedding light on the city’s future ESV and LER patterns. Furthermore, we established a connection between ESV and LER and subsequently redefined the ecological zones of Hohhot using Z-score standardization²⁶.

Fig. 2

Research Technology Roadmap. This image was made by Microsoft PowerPoint 2019 MSO (16.0.14326.20450).

Construction of landscape ecological risk index

Based on relevant literature, we divided Hohhot into a 2.5 km × 2.5 km square grid, which produced 2,941 evaluation units, with 2–5 times the average patch area used as a reference²⁷. We employed the ERI to assess the LER in Hohhot. Following the approach outlined by prior researchers^28,29,30 for constructing an ERI, our study computed various key indices. Specifically, we calculated the landscape fragmentation index, landscape disturbance index, landscape separation index, landscape fragility index, and landscape dominance index using Fragstats 4.2 (Table 2). The ERI for Hohhot was then calculated based on previous research³¹.

$$ERI = \sum\limits_{{i = 1}}^{n} {\frac{{A_{{ki}} }}{{A_{k} }}} \sqrt {E_{i} \times V_{i} }$$

(1)

Where: n is the number of landscape types; A_ki is the area of the i-landscape type in risk plot k; A_k is the total area of risk plot k; E_i is the landscape disturbance index; V_i is the landscape fragility index.

Table 2 Landscape pattern index and its meaning.

Assessment of ecosystem services value

We applied the unit area value equivalent factor method proposed by Xie³⁴. We selected the ratio of the 2022 farmland grain unit area yield in Hohhot (5473.5 kg/hm²) to the 2022 national farmland grain unit area yield (5805 kg/hm²) (0.94) as the correction factor. This factor was used to standardize the equivalent value for Hohhot City, resulting in 3202.11 Yuan/hm². Subsequently, we adjusted this value using the ESV equivalent per unit area table provided by Xie³⁴ to derive the ESV per unit area table for Hohhot City (Table 3). The formula for calculating the ESV is as follows:

$$ESV_{i} = \sum\nolimits_{{j = 1}}^{6} {A_{{ij}} } \times V_{j}$$

(2)

Where: ESV_i is the ecosystem service value of the i grid in the study area; A_ij is the area of the j land use type in the i grid; V_j is the ESV coefficient of the j land use type; ESV_i.e. is the value of i ecosystem service in the e grid; V_je is the value coefficient of ecosystem service function e for the j land use type.

Table 3 Value coefficients of various ecosystem services in Hohhot (CNY/hm²).

Value-risk ecological zoning quadrant

We applied Z-score standardization to the ESV and LER for Hohhot, which enabled the rezoning of Hohhot. In this study, Z-score normalization was applied separately to the ESV and LER. The normalized ESV and LER were then plotted in a two-dimensional coordinate system, with ESV on the x-axis and LER on the y-axis, dividing the space into four quadrants. These quadrants represent different ecological zoning types: the first quadrant indicates high ecosystem service value-high ecological risk zone (Zone I); the second quadrant represents low ecosystem service value-high ecological risk zone (Zone II); the third quadrant corresponds to low ecosystem service value-low ecological risk zone (Zone III); and the fourth quadrant signifies the high ecosystem value-low ecological risk zone (Zone IV). The Z-score normalization formula is as follows:

$$\begin{gathered} x = \frac{{x_{i} - \bar{x}}}{S} \hfill \\ \bar{x} = \frac{1}{n}\sum\limits_{{i = 1}}^{n} {x_{i} } \hfill \\ S = \sqrt {\frac{1}{n}\sum\nolimits_{{i = 1}}^{n} {(x - \bar{x})} ^{2} } \hfill \\ \end{gathered}$$

(3)

Where: x is the standardized value of the ecosystem service and landscape ecological risk index for Hohhot; x_i is ESV and ERI for the i-evaluation unit; $\overline{x}$ is the mean ESV and LER for Hohhot; S is the Standard deviation of Hohhot as a whole; n is the number of evaluation units.

Ecological zoning predictions based on the PLUS model

In this study, we selected two land use simulation models: the PLUS model proposed by HPSCIL@CUG Laboratory and the GeoSOS-FLUS model developed by Sun Yat-sen University. The PLUS model combines the Land Expansion Analysis Strategy and the Cellular Automata model, incorporating the Multi-class Stochastic Seed of Plaques for predicting and deducing land use patterns within the study area. On the other hand, the GeoSOS-FLUS model utilizes an adaptive inertial competition mechanism based on a roulette wheel, enhancing the traditional CA model with the help of Artificial Neural Networks (ANN) to predict land use changes. This approach allows for highly accurate simulation of real-world land use distribution. In this study, the accuracy of the model was assessed using the Kappa coefficient consistency test³⁵ The Kappa value falls within the range of 0 to 1. It is typically interpreted based on the following criteria: very low consistency (0–0.2), medium consistency (0.2–0.4), fair consistency (0.4–0.6), good consistency (0.6–0.8), and nearly perfect consistency (0.8–1).

Utilizing land use data from 2000 to 2010 as a foundation, we conducted simulations of Hohhot’s land use distribution for the year 2020 using both the PLUS model and the GeoSOS-FLUS model. The results (Fig. 3) indicated that the land use distribution map generated by the PLUS model achieved a Kappa coefficient of 0.86, with an overall accuracy of 0.9. In contrast, the GeoSOS-FLUS model yielded a Kappa coefficient of 0.84 with an overall accuracy of 0.89. While the discrepancy in Kappa coefficients between the two models is minimal, an examination of Fig. 3 demonstrates that the PLUS model offered a more refined simulation of Hohhot City’s land use distribution pattern compared to the GeoSOS-FLUS model. As a result, this paper selected the PLUS model for future predictions of land use distribution patterns.

To explore land use changes in Hohhot under various development objectives, this study, following the selection of an appropriate land use simulation model, established four scenarios: Natural development, Ecological protection, Arable protection, and urban development. These scenarios were devised in alignment with Hohhot’s comprehensive land use plan. The study projected the spatial distribution of land use in Hohhot for the year 2040, taking into account the city’s land use and economic development. Within these four scenarios: In the natural development scenario, each land type will continue its existing development pattern without modification. In the arable land protection scenario, restrictions will be imposed on the conversion of arable land into other land types. In the ecological protection scenario, forests, grasslands, and water areas will be safeguarded against conversion into other land types. In the urban development scenario, the conversion of other land types into construction land will take precedence.

Fig. 3

Actual and simulated land use differences in Hohhot in 2020. This image was created with Adobe Illustrator 2022 software (Adobe, San Jose, CA, USA) and ArcGIS 10.8 software (Esri, Redlands, CA, USA) using the results produced by PLUS model and FLUS model, and the data from Resource and Environment Science and Data Center (https://www.resdc.cn/).

Results

Analysis of land use changes

Spatial distribution pattern of land use in Hohhot from 2000 to 2020

Figure 4 reveals significant spatial distribution differences in land use types across Hohhot City from 2000 to 2020. Arable land had the widest distribution, primarily concentrated in the northern part of Wuchuan County, Tumote Left County, and Tuoketuo County. Grasslands were mainly located in the southern part of Wuchuan County and Qingshuihe County. Water areas were sparse, primarily concentrated in the western part of Tumote Left County, with a relatively stable pattern. Over the 20 years, the expansion of construction land was particularly prominent, radiating outward from Yuquan District to Saihan District, Huimin District, and Xincheng District, with a significant increase in area. This trend highlights the strong impact of rapid urbanization on land use patterns. Meanwhile, the expansion of construction land led to a reduction in grassland and arable land areas, indicating substantial pressure on ecological and agricultural land due to urbanization. In the future, it is essential to further balance land resource development and ecological protection to achieve regional sustainable development.

Fig. 4

Spatial pattern of land use in Hohhot from 2000 to 2020. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA). The data was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Land use pattern simulation and multi-scenario prediction in Hohhot

Figure 5 illustrates that unutilized land in Hohhot City for the year 2040 has decreased in comparison to 2020 across the four development scenarios. In the urban development scenario, the area of construction land increased by 91.2 km² compared to 2020, while arable land, grassland, and water areas decreased by 86.56 km², 11.43 km², and 2.75 km², respectively. This suggested that urban development encroaches on arable land, grassland, and water areas during the expansion of construction land. Under the arable land protection scenario, the arable land area was 6,294.81 km², marking a 0.7% increase from 2020. The spatial distribution of arable land expanded across the study area compared to Hohhot City’s 2020 land distribution pattern. In the natural development scenario, compared to the area of each category in Hohhot City in 2020, the area of construction land and forest land increased, while arable land, grassland, and water areas decreased. The most substantial reduction occurred in arable land, which decreased by 73.02 km², This indicated that the increase in construction land and forest land under the natural scenario was primarily due to the decrease in arable land. In terms of distribution, construction land expanded mainly in the north and central parts of the study area, while forest land increased primarily in the central and western areas. Under the ecological protection scenario, ecological land categories such as grassland, forest land, and water areas increased compared to 2020, indicating the effectiveness of ecological protection measures.

Fig. 5

Predictions of land use types for Hohhot in 2040 under Urban, Natural development and arable land protection, Ecological protection scenarios. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using the results produced by PLUS model. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Characterizing spatial differentiation in landscape ecological risk

Quantitative characteristics of landscape risk

Using the natural break point method, the LER of Hohhot was categorized into five grades: Very low risk (LER < 0.0076); Low risk (0.0076 < LER < 0.0098); Medium risk (0.0098 < LER < 0.013); High risk (0.013 < LER < 0.017); Very high risk (LER > 0.017). Subsequently, we conducted a detailed analysis of the ecological risk area across these different landscape grades in Hohhot (Table 4). The results revealed that the area with the very low ecological risk exhibited a consistent upward trend from 2000 to 2020. In 2020, the area with the very low ecological risk reached its peak, covering 31.36% of Hohhot’s total land area. Conversely, in 2000, the area with the very low ecological risk was at its lowest, accounting for 24.43%. The area with a low ecological risk decreased by 23.1% over the course of 20 years, measuring 7681.24 km² in 2000, 6117.87 km² in 2010, and 5906.63 km² in 2020. The area with a medium ecological risk showed a decreasing trend. Meanwhile, the area with high ecological risk increased by 22.2% over 20 years. The very high ecological risk area exhibited a fluctuating upward trend, growing by 201.36 km² during this period.

When considering the ecological risk class structure, Hohhot City was primarily characterized by the prevalence of the very low ecological risk class, the low ecological risk class, and the medium ecological risk class in the three periods spanning from 2000 to 2020, covering 94.11%, 91.11%, and 90.95% of the city’s area, respectively. Conversely, the area occupied by high risk ecological classes and very high risk ecological classes in these three periods amounted to just 5.89%, 8.89%, and 8.05% of the city’s total area. Notably, the proportion of the area with a higher ecological risk grade was even smaller, indicating a significant enhancement of ecological security in Hohhot.

Table 4 Area and proportion of landscape ecological risk level in Hohhot from 2000 to 2020.

Based on the ecological risk level threshold interval division of Hohhot City, the spatial mapping information for LER during three periods (Fig. 6) reveals distinct patterns. The LER in Hohhot City was characterized by a widespread and contiguous distribution of the very low and low ecological risk regions. The distribution of medium ecological risk regions was more dispersed, exhibiting characteristics of peripheral distribution around the very low and low risk regions, indicating spatial transitions. High ecological risk regions and very high ecological risk regions were closely linked, demonstrating sporadic and centralized spatial distribution characteristics on the map. Analyzing the specific land use distribution data across three periods, the water areas and construction land classes predominantly exhibited the very high ecological risk levels and high ecological risk levels, respectively. Meanwhile, the forest land and grassland classes primarily manifested the very low ecological risk levels and low ecological risk levels during these three periods. As for the arable land class, it was characterized by a predominance of medium ecological risk levels.

Fig. 6

Spatial map of landscape ecological risk level in Hohhot from 2000 to 2020. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using the results produced by Fragstats 4.2. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Quantitative characterization of landscape risk in future scenarios

Applying the same thresholds as those in Sect. 4.2.1, Figure 7 illustrates a noteworthy shift in the ecological risk class structure of Hohhot’s landscape in 2040 compared to that of 2020 across the four scenarios. Specifically, in the urban development scenario, the very high risk area experienced the most substantial increase, growing by 4.14%, expanding from 320.27 km² in 2020 to 1030.96 km² in 2040, with a pronounced enlargement in the high risk area. In contrast, in the arable land protection scenario, the very low risk area was the smallest, while the low risk area was the most extensive among the scenarios.

Fig. 7

Comparison of the proportion of various landscape ecological risk areas in Hohhot in 2020 and 2040.

Figure 8 offers a spatial overview of Hohhot in 2040 within the contexts of urban development, natural development, and ecological protection scenarios. In these scenarios, high ecological risk was primarily concentrated around the very high risk, particularly in the urban area and its adjacent counties. Notable examples included the southern part of Tumote Left County and Tuoketuo County, characterized by dominant land use types such as construction land, arable land, and water areas. Low ecological risk tended to surround the regions with the very low ecological risk. The low ecological risk and very low ecological risk were primarily found in Wuchuan County, Xincheng District, Huimin District, and the southern part of Tumote Left County. In these regions, the primary land types consisted of forested land, grassland, and construction land. The medium ecological risk was situated at the border of high ecological risk and low ecological risk, presenting a fragmented distribution pattern, primarily in the southeastern part of the study area, including Helingeer County, where arable land and grassland were predominant land types. The arable land protection scenario brought about substantial changes in the distribution and area of LER. For example, medium ecological risk emerged in the northern and central-western regions of the study area, while high ecological risk and very high ecological risk exhibited sporadic and concentrated patterns.

Fig. 8

Spatial distribution of landscape ecological risk index in Hohhot under multiple scenarios in 2040. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using the results produced by Fragstats 4.2. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Characteristics of spatial differentiation in the value of ecosystem services

Spatial-temporal characteristics of ecosystem service values

Based on Table 5, the ESV of Hohhot City in 2000, 2010, and 2020 amounted to 62.08 × 10⁹ yuan, 62.19 × 10⁹ yuan, and 62.06 × 10⁹ yuan, respectively, reflecting minimal changes over this period. When examining the value of land resources, it became evident that grassland and forest ecosystems made the most significant contributions to the ESV, collectively accounting for over 65% of the total. The ESV of grassland in 2000, 2010, and 2020 stood at 24.94 × 10⁹ yuan, 25.18 × 10⁹ yuan, and 24.2 × 10⁹ yuan, respectively, representing 40%, 40%, and 39% of the overall ESV. Forest ESV for these years amounted to 16 × 10⁹ yuan, 15.66 × 10⁹ yuan, and 16.65 × 10⁹ yuan, constituting 26%, 25%, and 27% of the ESV, respectively. Water areas and arable land ecosystems also played a substantial role in Hohhot City’s ESV. Over the same period, their ESV was 12.67 × 10⁹ yuan, 12.89 × 10⁹ yuan, 13.09 × 10⁹ yuan, 8.37 × 10⁹ yuan, 8.39 × 109 yuan, and 8.03 × 10⁹ yuan, with contributions of 26%, 25%, 27%, 13%, 13%, and 13%, respectively. In summary, grassland, water areas, forests, and arable land constituted the primary components of Hohhot City’s ESV, and managing land use transitions carefully could lead to ESV improvements in Hohhot City.

Table 5 Value and proportion of ecosystem services by land use type in Hohhot from 2000 to 2020.

The study employed the natural breakpoint method, considering the land resource distribution pattern within the city, to categorize ESV into five levels: lowest value (ESV < 0.015 × 10⁹); lower value (0.015 × 10⁹ < ESV < 0.023 × 10⁹); medium value (0.023 × 10⁹ < ESV < 0.031 × 10⁹); higher value (0.031 × 10⁹ < ESV < 0.057 × 10⁹); and highest value (0.057 × 10⁹ < ESV < 0.09 × 10⁹). The result was a series of ESV spatial pattern distribution maps for different periods (Fig. 9). Looking at the specifics, the highest value regions were concentrated around the Tumote Left County, driven by the significant role of river systems in water conservation, water supply, and their high ecological importance, which translated to greater value and functionality. Regions with higher value were primarily situated in the northern part of Hohhot City, predominantly consisting of forest ecosystems dominated by Chinese pine and juniper. These ecosystems offered stable high productivity and a wide array of ecological service functions, resulting in a relatively high ESV. The medium value regions exhibited a widespread and somewhat dispersed distribution across Hohhot City, displaying characteristics of peripheral distribution around the highest value regions. This pattern was primarily due to the substantial proportion of arable land ESV within the medium-value regions. Lowest value and lower value regions also had a relatively extensive distribution. Concerning land resource structure, forest, and grassland ecosystems were mainly categorized as medium value or higher.

Fig. 9

Spatial map of ecosystem service value levels in Hohhot from 2000 to 2020. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Changes in the value structure of ecosystem services in the future

The ESV (Fig. 10) exhibits higher concentrations in the north and lower values in the south, with medium value regions encircling higher value regions. These regions were mainly situated in the southern part of Wuchuan County and the northern part of Tumote Left County, with a smaller portion showing a mosaic-like distribution in the southeastern part of Helingeer County. These regions were known for their rich biological resources, such as grassland, forest land, and arable land. Highest value regions were sporadic, primarily around water areas, such as the Hasuhai Lake in the Tumote Left County, lower value and lowest value regions displayed fragmented distributions throughout the entire study area. When comparing the spatial distribution maps of ESV among the four scenarios, it was observed that the area of lowest value regions decreased in the scenarios prioritizing ecological protection, Natural development, and arable protection compared to the urban development scenario. Concurrently, there was a notable increase in the regions of lower value, medium value, and higher value regions, particularly in the western part of Saihan District, the northern section of Wuchuan County, and the central part of Yuquan District. Notably, the ecological protection scenario exhibited the most significant increase.

Fig. 10

Spatial distribution of ecosystem service value in Hohhot under multiple scenarios in 2040. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Construction of ecological space zoning

Current eco-spatial zoning construction

Z-score standardization was employed to calculate ESV and LER, resulting in the generation of the ecological zoning quadrant distribution map of Hohhot (see Fig. 11), the spatial distribution mapping of ecological zoning (Fig. 12), and the statistical values of ecological zoning (Table 6):

Fig. 11

Distribution of ecological zones in Hohhot from 2000 to 2020.

Hohhot City was primarily characterized by the Zone IV, which accounted for 41.49%, 42.61%, and 41.97% of Hohhot City’s total area in 2000, 2010, and 2020, respectively. This zone exhibited a fluctuating upward trend in the area from 2000 to 2020 and was widely distributed throughout Hohhot City, with its main concentration in the southern part of Wuchuan County, the northern part of Tumote Left County, and the northern part of Xincheng District. The primary land-use types in these regions were arable land and grassland.

The Zone I had the smallest area, displaying a decreasing trend with 13.3 km², 0.38 km², and 4.94 km² in the three periods, respectively. It exhibited a high degree of overlap with the river system and was concentrated in the Hasuhai Lake in the middle of the Tumote Left County. The area of the Zone II displayed a decreasing trend, decreasing by 1.24% from 2000 to 2020. It exhibited a mosaic distribution, was surrounded by the Zone III, and was mainly concentrated in the area where construction land and arable land were interspersed. The area of the Zone III displayed an increasing trend, with 5773.66 km², 5894.09 km², and 5911.57 km² in the three periods, respectively. The primary land-use type in this ecological zone was arable land, which was interspersed and distributed within the Zone II and Zone IV.

In summary, the Zone IV in Hohhot City was predominant and exhibited an increasing trend in the area. Meanwhile, the Zone I and the Zone II showed a decreasing trend. The area of the Zone III was on the rise, indicating an overall improvement in the ecological zone structure of Hohhot.

Fig. 12

Spatial distribution map of ecological zones in Hohhot from 2000 to 2020. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Table 6 Statistics of ecological divisions of Hohhot City in 2000–2020.

Construction of ecological space zoning under multiple scenarios in the future

The Z-score standardized calculation of ESV and LER under the future scenario was employed to generate the future ecological zoning quadrant distribution map of Hohhot (Fig. 13), the spatial distribution mapping of ecological zoning (Fig. 14), and the statistical values of ecological zoning (Table 7).

Fig. 13

Distribution of future multi-scenario ecological zones in Hohhot.

Fig. 14

Spatial distribution map of future multi-scenario ecological zones in Hohhot. This image was made in ArcGIS 10.8 software (Esri, Redlands, CA, USA) using. The original boundary was obtained from Resource and Environmental Science Data Platform (https://www.resdc.cn/).

Under the urban development scenario, in comparison to 2020, the area of the Zone I and the Zone II exhibited a decreasing trend. Zone II experienced the most significant reduction, with a decrease of 257.2 km². On the other hand, the area of the Zone III and the Zone IV demonstrated an increasing trend. Zone III and Zone IV both showed an upward trend, with an increment of 264.8 km². Under the natural development scenario, when compared to the year 2020, the regions of the Zone I, Zone II, and Zone IV exhibited a decreasing trend. Collectively, these zones experienced a total decrease of 1331.64 km². Conversely, the Zone III showed an increasing trend, suggesting that ecological land may have been protected to some extent during the natural development process. Under the ecological protection scenario, in comparison to the year 2020, the area of Zone I decreased. However, it remained the largest among all scenarios. The distribution of Zone I was concentrated primarily in the Hasuhai Lake. This suggested that ecological protection policies have been effective in safeguarding ecological land in this area, while the remaining regions had experienced relatively minor changes in the area. Under the arable land protection scenario, in contrast to the year 2020, Zone III witnessed a substantial increase in area. However, the other ecological sub-zones experienced a decline in area. This suggested that the process of arable protection might have adverse effects on other ecosystems and the environment.

Table 7 Statistics of ecological zoning in Hohhot under multiple scenarios in the future.

Discussion

Analysis of spatiotemporal changes of land use dynamics in Hohhot

This study analyzed land use changes in Hohhot from 2000 to 2020 using a land use transition matrix. By examining the areas of land converted into and out of various categories, it was observed that the areas of grassland, arable land, and unused land decreased, while those of construction land, water areas, and forestland increased. Among these, arable land exhibited the largest net decrease due to its conversion into grassland, forestland, and construction land. This transformation could be attributed to two main factors: first, large-scale environmental projects promoting afforestation and grass planting, implemented between 1990 and 2020, contributed to the reduction of arable land; second, Hohhot’s rapid economic growth and population expansion in recent years led to urban sprawl, encroaching on arable land. In addition to arable land, other land categories also decreased due to conversion into construction land. If this trend continues, it may adversely affect Hohhot’s future development.

From a future perspective, analyzing land use in Hohhot revealed the critical role of urban development planning in regulating land use changes. For instance, under an arable land protection scenario, arable land increased, indicating that implementing arable land protection policies and strictly regulating encroachment by other land categories effectively safeguarded arable. In contrast, under a natural development scenario, the areas of construction land and forestland increased, while arable land decreased. This reduction was primarily due to the afforestation policies implemented between 2006 and 2020 and the expansion of construction land driven by economic growth and population increase. As the natural development scenario aligned with Hohhot’s past land use trends, continued afforestation policies would have resulted in further conversion of arable land into forestland.

In conclusion, land use change was a complex issue influenced by social, economic, environmental, and policy factors. Future land use planning in Hohhot should have emphasized sustainable land utilization, minimized encroachment on other land types—especially arable land and natural ecological land—during urban expansion. A balanced allocation of land use structures is essential to maximize economic, social, and ecological benefits. Additionally, it is crucial to strengthen land use planning and management to ensure sustainable and healthy urban development through scientifically sound land use strategies and stringent oversight.

Fig. 15

Land use transfer matrix in Hohhot City from 2000 to 2020 (km²).

The impact of land use change on landscape ecological risks

The spatial structure of ecological risk and changes in the spatial distribution pattern of land use and land area exhibit consistency and synchronization²⁶. Alterations in the spatial distribution pattern of land use and land area will impact the NP and their CA, subsequently influencing the LER.

This study initially analyzed the spatial distribution of LER in Hohhot from 2000 to 2020. The results revealed that Hohhot’s landscape ecological safety pattern was primarily characterized by the very low ecological risk, the low ecological risk, and the medium ecological risk. These ecological risk classes exhibited a fluctuating decrease in area over time, while the high ecological risk and very high ecological risk classes displayed a fluctuating increase in area. This suggested the possibility of an increase in the LER in Hohhot, which could have posed a potential challenge in optimizing the landscape ecological security pattern in the region. In the future LER distribution, the urban development scenario restricted the conversion of construction land, making it less likely to transition compared to the 2020, nature development, arable land protection, and ecological protection scenarios. This resulted in a continuous expansion of construction land and encroachment on other land classes, causing fragmentation of these remaining land class patches. Consequently, there was an increase in the NP and a decrease in the CA, leading to a higher degree of landscape loss and reduced resistance to disturbances. This transformation resulted in a decrease in the area of very low risk and low risk, with a shift towards medium risk, high risk, and very high risk. In both the ecological protection and natural development scenarios, the emphasis on safeguarding Hohhot’s ecosystems prevented the encroachment of other land classes, such as grassland, water areas, and forestland, by restricting the expansion of construction land. However, due to economic and population growth, the expansion of construction land was unavoidable. As a result, the expansion of construction land encroached on arable land, leading to the fragmentation of arable land patches. This resulted in an increase in the NP, a decrease in CA, and an escalation of landscape loss. Arable land, one of Hohhot’s primary land types, had a significant share of the total area, second only to grasslands. The increase in arable land landscape loss contributed to the expansion of higher risk within Hohhot. Additionally, when encroaching on other land categories, the conversion to construction land leads to fragmentation of its patches, increased landscape separation, and reduced stability. Under the arable land protection scenario, arable land was strictly protected from conversion, and other land categories were prohibited from encroaching on arable land when expanding. Analyzing Hohhot’s land transfer matrix from 2000 to 2020 (Fig. 15), it was evident that arable land was mainly converted into grassland and construction land, while grassland was partially converted into arable land. Therefore, the decrease in arable land area was mainly due to the expansion of construction land. However, the expansion of construction land was not allowed to encroach on arable land in this scenario, resulting in minimal construction land area. Consequently, there was an increase in landscape separateness and loss indexes, and the lowest risk transitioned into lower risk. The transfer probabilities of the other land categories followed Hohhot City’s 2020 land transfer patterns, resulting in relatively minor changes in the NP and patch area CA, with minimal alteration in risk levels. A comprehensive analysis of the above findings underscored the pivotal role that the city’s development decisions played in shaping the future LER regions of Hohhot City. When formulating Hohhot City’s future development plan, it is imperative to prioritize the protection of ecological lands, including grasslands, forest lands, and areas of water, and to place a strong emphasis on safeguarding arable land, particularly fundamental arable land. This could be achieved through strategic adjustments to the land-use structure, ensuring comprehensive control over the distribution patterns of land use and facilitating the transformation of land categories. By effectively managing the spatial structure of LER, we could pave the way for the sustainable development of Hohhot City.

Impacts of land use on ecological zoning

According to the research results, in the ecological zoning of Hohhot City for the years 2000, 2010, and 2020, Zone IV predominated, while Zone I, Zone II, and Zone III did not show significant dominance. This phenomenon could be attributed to several factors: first, the dominant land use types in Zone IV were forests and grasslands, which had a high ESV. From 2000 to 2020, grasslands and forest ecosystems made a significant contribution to the ESV of Hohhot City (65%). These ecosystems played crucial roles in stabilizing soil, conserving water, providing habitats, and regulating climate, which enhanced their ecosystem service value. Additionally, these ecosystems were relatively stable and faced lower ecological risks. The natural conditions of Hohhot City might have been more conducive to the maintenance and development of ecosystems in Zone IV, particularly in terms of terrain and climate, which provided a favorable foundation for the growth of forests and grasslands, thereby maintaining higher ESV and low ecological risks.

In contrast, Zone I overlapped heavily with river systems and was concentrated in the central area of Hasi Sea in Tumote Left County. This unique ecological environment made the ecosystems in this region relatively fragile, vulnerable to human activities and natural disturbances, resulting in higher ecological risks. For example, water ecosystems were subject to pollution and changes in water resources, which, despite their high ESV, they faced substantial ecological risks. The area of Zone I remained small at all three time points, measuring 13.3 km², 0.38 km², and 4.94 km², and its distribution is relatively concentrated in specific regions, limiting its proportion within the city’s overall ecosystem and preventing it from becoming a dominant zone.

Zone II primarily consisted of a mix of construction land and arable land/grassland. The expansion of construction land often leads to the fragmentation of arable land, increasing landscape fragmentation and reducing ecosystem stability and service value. For instance, during urban expansion, the encroachment of construction land on arable land resulted in the reduction of arable land patches, damaging ecosystem integrity and lowering service value, thus contributing to high ecological risks. Furthermore, Hohhot City’s policies of converting arable land to forest and grassland had accelerated the occupation of arable land by grasslands and forests, further intensifying arable land fragmentation, making the ecosystem in this area more vulnerable, and leading to continued degradation in both ecosystem service value and ecological security.

Zone III, primarily consisting of arable land, although important in the ecosystem for food production, had a lower ESV compared to forests and grasslands. For example, arable land mainly provided ecosystem services related to food production, but contributes less to climate regulation and biodiversity maintenance, resulting in a lower overall ESV in this zone. While ecological risks were low in Zone III, urbanization, driven by increased construction and forest land, had reduced the area of arable land. The conversion of arable land due to afforestation policies and urban expansion highlighted the vulnerability of arable land to urban development and policy changes, limiting its stability and, consequently, its dominance within the ecosystem.

In conclusion, the characteristics and advantages of Hohhot City’s different ecological zones were the result of a complex interplay of natural and human factors. Future urban planning should take these factors into account and implemented scientifically-based planning, management, and protection measures to ensure sustainable urban ecological development, enhance urban ecological resilience, and improve the quality of life for residents.

Applicability of integrating ESV-LER to delineate ecological zones

ESV and LER are two critical dimensions in ecosystem studies, each reflecting the internal stability and external contributions of ecosystems. LER focuses on evaluating the spatial structure of ecosystems and their sensitivity to human disturbances, mainly addressing the internal stability and resistance of ecosystems to disturbances. In contrast, ESV reflects the services provided by ecosystems composed of various land-use types, focusing on the contributions and impacts of ecosystems on the external environment. The combination of these two dimensions offers a more comprehensive understanding of both the internal and external dynamics of regional ecosystems, facilitating an in-depth analysis of the spatiotemporal evolution of land use, ecosystem services, and ecological risks, thus providing a more comprehensive ecological security analysis framework^23,36. Previous studies predominantly focused on static or single-indicator frameworks. For example, the studies by Li et al.³⁷, and Liu et al.³⁸, and Yang et al.³⁹ mainly concentrated on one dimension of ecosystem services or ecological risk, without fully considering the interaction and synergies between ecosystem services and landscape ecological risk. While these studies provided valuable insights into the ecological characteristics of specific regions, they often overlooked the dynamic relationship between ecosystem services and landscape ecological risk. As a result, their applications in ecological zoning and risk assessment are limited, often providing localized or static analysis results, lacking a comprehensive understanding of ecosystem health and regional ecological security. Additionally, existing ecological zoning studies are primarily focused on static assessments, lacking predictions of future changes. For instance, Shen et al.⁴⁰ and Gan et al.⁴¹ conducted ecological zoning for Fangchenggang and Shanghai, respectively, from a static perspective, which may lead to a conservative or partial understanding of regional ecological security. In contrast, this study used dynamic simulation of future scenarios, allowing for a better capture of the changing trends between ecological risk and ecosystem services, providing more enriched decision support for future ecological management and policy formulation.

In our studies, we integrated the two dimensions of ESV and LER to construct ecological zones and employed the PLUS model to simulate multiple future scenarios. This approach successfully overcame the limitations of traditional ecological zoning in the temporal dimension and provided a dynamic, multi-indicator analysis perspective, offering a novel approach to ecological security theory.

Ecological zoning management and control scheme

Hohhot could be categorized into four types of reserves based on the ESV and LER of the grid cells, as shown in Fig. 16: ecological restoration reserve, ecologically rich reserve, ecologically balanced reserve, and ecologically challenged reserve. These categories were described as follows:

Ecological Restoration Reserve: This type of area exhibited both a high ESV and a high LER level. It served as a vital area for guaranteeing ecological service functions and as a top priority for habitat restoration within Hohhot City. Typically, this category was characterized by ecosystems such as water areas. To ensure the continued enhancement of ecological service functions within the Ecological Restoration Reserve, future development planning should incorporate measures to minimize human activity interference, gradually enhance landscape integrity, reduce habitat fragmentation, and adopt an ecological management approach that combined natural and artificial restoration methods. The implementation of these measures will effectively maintain and enhance the ecological service functions of the Ecological Restoration Reserve, thereby fortifying Hohhot City’s ecological security.

Ecologically Rich Protected regions: This category of regions exhibited a low ESV and a high LER level, representing the quintessential ecologically fragile zones within Hohhot. It was primarily characterized by a mosaic of construction land and arable grassland, with fragile ecosystems, a high degree of landscape fragility, and limited internal stability. This phenomenon could be attributed to two main factors. Firstly, the encroachment of construction land into arable land during urban expansion resulted in fragmented patches of arable land, an increased NP, reduced CA, and heightened landscape degradation. Secondly, the policy of converting arable land into forests and grasslands in Hohhot City led to the encroachment of grasslands and forests into arable land, further exacerbating the fragmentation of arable land. As depicted in Fig. 14, the area of ecologically rich protected regions in the 2040 arable land protection scenario within Hohhot City was the smallest compared to other scenarios. This finding validated our analysis, which identified arable land encroachment as the primary cause for the expansion of ecologically rich protected regions. Moreover, the implementation of arable protection policies could effectively reduce the extent of ecologically rich protected regions. Consequently, future planning for Hohhot City should prioritize the protection of arable land and strive for a balanced transformation of land types concerning arable land.

Ecologically Balanced Protected Area: This category of area was characterized by a low ESV and a LER level. It played a pivotal role in the comprehensive improvement and overall quality enhancement of Hohhot City’s ecological landscape. The predominant land use type within this area is arable land, and its extent showed a consistent increase from 2000 to 2020. Considering Hohhot City’s land use transfer matrix and future multi-scenario ecological zoning, we observed that changes in arable land use are the primary drivers affected the size of ecologically balanced protected regions. These regions expanded with the increase in arable land, emphasized the importance of preserving and enhancing the ecological quality of arable land for elevating the ESV and reducing LER in the future. In Hohhot’s future planning, it will be crucial to protect and enhance the ecological environment of arable land, reinforce Arable protection and management, and ensure both the quantity and quality of arable land. Additionally, efforts should focus on the maintenance and restoration of arable land to improve its fertility and water utilization, thus safeguarding its production capacity.

Ecologically Balanced Protected Area: This region was characterized by high ESV, low LER levels, and favorable ecological resource conditions, made it a critical source of ecological security for Hohhot City, particularly due to the prevalence of forest and grassland ecosystems. Tables 3 and 4 revealed that, under the arable land protection scenario, the region of ecological challenge protection zones is the largest, followed by Urban development and ecological conservation scenarios. A comparison with the year 2020 indicated an increase in the region of ecological challenge protection zones under urban development and a slight decrease under the ecological protection scenario. This trend could be attributed to the implementation of ecological protection policies in the ecological protection scenario, aimed at safeguarding ecological lands such as forests, grasslands, and water areas from encroachment by other land types. However, the conversion of other land types resulted in fragmented portions connected to ecological lands, leading to an increase in patch quantity and landscape fragmentation. In contrast, under the scenarios of urban development and arable protection, protecting only one land type from encroachment by others preserved the integrity of that land type. This outcome emphasized the significance of selecting appropriate ecological protection areas, as implementing Ecological protection measures in an unreasonable manner could result in adverse consequences, despite the importance of ecological conservation. In summary, the interplay between the shifts in the area of each ecological subregion in Hohhot from 2000 to 2020 and the variations in the area of ecological subregions across different future scenarios underscored the significant influence of urban planning on the natural environment. To safeguard the ecological environment and uphold ecosystem equilibrium, it will be imperative to implement prudent management and protection measures. This will entail bolstering the management and protection efforts to guarantee the adequate preservation of the ecological functions within high-value ecological regions.

Fig. 16

Characteristics of ecological space zoning in Hohhot City.

Limitations and uncertainties

This study integrated remote sensing technology, geographic information modeling methodology, landscape ecology methodology, GIS grid analysis methodology, and land use simulation models. It combined the two dimensions of ES and LER to elucidate the spatial and temporal evolution characteristics of land use, ESV, LER, and ecological zoning in Hohhot City across various current and future scenarios. The study proposed a customized control plan for ecological zones in Hohhot City, enhanced the spatial representation of ecological resources and presented a novel approach for spatially identifying ecological functional zones within natural protected regions.

However, this study has several limitations. First, regarding land use simulation, even though we utilized the high-precision PLUS model, it relied on fixed transition rules for each land category when projecting future land use changes. These rules may change due to the influence of future policies and planning⁴². Furthermore, since relevant planning documents for the study area had not been published, we had to create future scenarios by referencing other research materials. Although the scenarios in this study did not align with national spatial planning, they provided valuable insights into the changes in land categories within Hohhot by 2040 under four scenarios: Arable protection, ecological protection, natural development, and urban Development. Consequently, the results of this paper can serve as a reference for future land spatial planning and sustainable development policies. In terms of ESV, our ESV assessment was closely tied to land use change patterns. In this study, land use types were classified into only six categories (arable land, forest land, grassland, water areas, construction land, and unused land), which may introduce some discrepancies between ESV assessment and actual conditions^43,44. Additionally, while the equivalent factor method was a more concise and effective approach, setting the service value coefficient of construction land to zero introduced a degree of uncertainty in the ESV assessment⁴⁵. Regarding the LER, the assignment of land classes significantly impacts the index and ecological zoning. In this study, when calculating the LER, land classes were scored based on empirical knowledge, leading to disparities between the research results and the actual landscape ecological risk values. Furthermore, in terms of ecological zoning, the current spatial identification of ecological functions using a value-risk synthesis approach is still an area of ongoing exploration. In addition, Among the natural factor data, the soil type data dated back to 1995, and the soil types may have changed due to human activities or natural processes. The research path was not yet well-established, and there were limited empirical cases to draw from.

In future research, we intend to address the following aspects to enhance the accuracy of our assessments: (1) Enhanced Land Use Data: prioritize the use of land use data with higher accuracy and finer granularity wherever possible. This will ensure a more precise representation of land use patterns within the study area. (2) Improved Land Use Simulation: enhance the land use simulation model to reduce prediction uncertainty. This improvement will lead to more reliable projections and a better understanding of future land use changes in the region. (3) Robust Future Scenarios: develop more reliable future scenarios in alignment with national spatial planning. These scenarios will serve as a solid foundation for policy planning, offering valuable insights into potential land use developments. (4) Detailed Land Use Types: incorporate more detailed land use categories and rational equivalent coefficient tables to provide a more accurate assessment of the ESV in Hohhot. (5) Advanced Ecological Zoning Techniques: refine and expand the technical methods used for ecological zoning. Our goal is to better support the organized advancement of national urban construction by providing improved ecological zoning strategies.

Conclusions

There exists a robust correlation between the ESV and LER. This research focused on Hohhot City as the study area. The study reclassified ecological zones and simulated their spatial distribution based on ESV and LER indices. The findings indicated the following: (1) Historical Trends (2000, 2010, 2020): LER in Hohhot City was predominantly characterized by very low, low, and medium risk levels. Very low risk and low risk areas were extensively distributed in large contiguous patches, while medium risk areas were more scattered. In contrast, high risk and very high risk areas were spatially sparse and clustered. This spatial distribution provided a regionalized framework for ecological risk management and supported the implementation of targeted ecological protection and risk mitigation measures. Regarding ESV, overall changes were minimal, with grassland and forest ecosystems contributing the most (65%) to ESV, showing relatively stable values. These findings offered critical insights for the sustainable use of land resources and the prioritization of ecological protection areas. Ecological zoning was dominated by Zone IV, while Zone I accounted for the smallest area. (2) Future Scenarios: Under different development scenarios, LER exhibited varying trends. In the urban development scenario, high risk areas showed the largest increase, highlighting the potential threats to ecological security posed by rapid urbanization. Conversely, in the arable land protection scenario, very low risk areas were the smallest, while low risk areas were the largest, indicating that protecting arable land could effectively reduce ecological risk. Regarding ESV, all four scenarios displayed a north-high, south-low spatial distribution pattern. The varying changes in ecological zone areas across scenarios demonstrate that policy interventions could effectively regulate ecological zone dynamics.

This study comprehensively evaluated the spatiotemporal trends of LER and ESV through the analysis of land use changes, considering both historical and current conditions. Based on a thorough assessment of the potential risks facing ecosystems and their provided values, we conducted ecological zoning and further predicted ecological zones under four different future scenarios using the PLUS model. This research offered practical solutions for urban land use planning and ecological protection policies. Not only was this study of significant importance for the sustainable development of Hohhot, but it also provided valuable insights for other cities facing similar ecological risks and urbanization pressures. Future research could further explore the dynamic changes in ecological risk and service value, incorporating additional ecological service indicators and risk factors to refine regional ecological planning models.